Medical signals are typically acquired using linear and uniform sampling and data acquisition. Known data conversion systems usually require sampling at two times the maximum frequency of the medical signals (according to the Nyquist—Shannon sampling theorem) to achieve accurate signal sampling. This may result in over-sampling and redundant data acquisition for low frequency portions of a signal, for example, such as rest portions of a heart activity signal (e.g., an ECG signal). Additionally, the over sampling and data conversion of medical signals may result in unwanted effects and inefficient use of a medical device and system as well as burdening of interface electronic through resource intensive data compression and transmission and additional signal delay. Known medical signal sampling systems uniformly process different portions of medical signals and utilize high sampling rate for different parts of a medical signal, irrespective of frequency and information content. A system according to invention principles addresses these deficiencies and related problems.